issues in anti-corruption law - university of notre …ndlaw/jleg/issue_articles/volume3… · ·...

272

ISSUES IN ANTI-CORRUPTION LAW:

DRAFTING IMPLEMENTING REGULATIONS FOR ANTI-CORRUPTION CONVENTIONS IN CENTRAL EUROPE

AND THE FORMER SOVIET UNION

Bryane Michael, Stockholm School of Economics

INTRODUCTION

For over ten years, organizations such as the Organization for Economic

Cooperation and Development, the Council of Europe, and the United Nations

have been helping developing countries adopt legal measures to fight

corruption. The various anti-corruption conventions signed by a wide number

of countries represent perhaps the most common element of such work.1 Yet,

the adoption of the anti-corruption conventions has had a questionable impact

on reducing corruption. Indeed, the limited empirical studies available point to

little, if any, correlation between the extent to which several Central and

Eastern European countries have adopted conventions against corruption and

reductions in perceived corruption levels in that country.2 Qualitative studies of

work on anti-corruption from the region also fail to find any significant

relationship between the adoption of legislation aimed at reducing corruption

and more corruption related detections and prosecutions.3While ratified by

national parliaments, these anti-corruption laws are not being implemented in

executive agencies most prone to corruption—particularly the traffic police, the

1. Each of these organizations have promulgated numerous conventions introducing provisions with the aim of reducing corruption. For example, the Council of Europe has promulgated at least 4 conventions aimed at fighting corruption -- a criminal law convention, Criminal Law Convention on Corruption, January 27, 1999, ETS No. 173, a civil law convention, Civil Law Convention on Corruption of 1999, ETS No. 174, a convention on international legal assistance in criminal matters, Convention on the Legal Status of Migrant Workers , ETS No. 93, and a convention on the confiscation of the proceeds of crimes such as corruption, Convention on Laundering, Search, Seizure and Confiscation of the Proceeds from Crime and on the Financing of Terrorism, ETS No. 198. The UN and OECD also have encouraged the adoption of a number of conventions and recommendations. A detailed discussion of these conventions would take this paper well outside its main argument and I refer the reader to Professor Henning’s paper for a background on these conventions. Peter Henning, Public Corruption: A Comparative Analysis of International Corruption Conventions and United States, 18 ARIZ. J. INT’L & COMP. L. 793 (2001).

2. For one such study, see Franklin Steves & Alan Rousso, Anti-Corruption Programmes in Post-Communist Transition Countries and Changes in the Business Environment, 1999-2002 (Eur. Bank for Reconstruction & Dev., Working Paper 85, 2003), available at http://www.ebrd.com/pubs/econo/wp0085.pdf.

3. See generally JAMES ANDERSON AND CHERYL GRAY, WORLD BANK, ANTI-CORRUPTION IN

TRANSITION 3: WHO IS SUCCEEDING . . . AND WHY? (2006), available at http://siteresources.worldbank.org/INTECA/Resources/ACT3.pdf; Frank Emmert, Administrative and Court Reform in Central and Eastern Europe, 9 EURO. L. J. 3 (2003).

2010] Issues in Anti-Corruption Law 273

police, customs, and tax inspectors.

Using a law and economics approach to anti-corruption regulation, this

paper seeks to provide an answer to the following question: ―how should

executive agencies in many Central European and Former Soviet countries

write anti-corruption regulations?‖ Executive agencies should write

regulations, using economic theory as a guide, such that the social benefits of

anti-corruption regulation outweigh the social costs. The first section of this

paper will address the importance of anti-corruption regulation—showing why

anti-corruption legal work has not yet succeeded—and the reasons why the

current ―let’s regulate‖ approach may be ill-conceived, because such an

approach ignores the costs and benefits of regulation. The second section deals

with the assignment of liability for corruption offences—showing how the

allocation of liability should follow the cost-benefit principle applied in this

paper. The third part of the paper looks at the financing of anti-corruption

work, showing how to provide incentives to encourage corruption fighters to

work harder and whistleblowers to blow the whistle. The fourth section looks

at the issue of jurisdiction, showing how the assignment of the prosecution of

non-criminal corruption cases to various government agencies can speed up

processing times and decrease processing costs. The fifth section provides the

rationale for using regulatory instruments as a strong complement to the

legislative approach (namely just implementing more detailed anti-corruption

acts). The final section recaps the argument that anti-corruption regulations that

implement the international anti-corruption conventions and national

legislation should generate positive social welfare (namely, their benefits

should exceed their costs). Given the newness of the topic (and space

limitations), the paper provides a particularly polemic approach to the topic—

leaving a more balanced and critical review of the topics for future work.

THE IMPORTANCE OF ANTI-CORRUPTION REGULATION

The question of anti-corruption regulation raises immense interest in

Figure 1: Adoption of Anti-Corruption Conventions and Control

of Corruption

0

0.2

0.4

0.6

0.8

1

0 10 20 30 40 50 60 70 80 90

Control of Corruption (percentile)

Ind

ex

of

Ad

op

tio

n o

f

Co

nv

en

tio

ns

(h

igh

er

nu

mb

ers

im

ply

fu

lle

r

ad

op

tio

n)

Source: Kaufmann et al. (2005) and Steves and Alan Rousso (2003)

274 Journal of Legislation [Vol. 36:272

policymaking circles because the current approach is not working well for most

of Eastern Europe. Figure 1 plots a variable which Steves and Rousso use to

measure the extent to which several Central and Eastern European countries

have adopted anti-corruption conventions against the extent to which ―public

power is exercised for private gain, including both petty and grand forms of

corruption.‖4 As shown in comparison (by a widely cited variable which

Kaufmann et al. refer to as ―control of corruption‖)—which is a rank list of

countries according to the perceived effectiveness of executive agencies in

fighting corruption—this figure shows no correlation across countries between

the extent to which a country adopts anti-corruption conventions and the

ability of its public administrators to fight against corruption.5 Indeed, if a

general trend is discernible in the data, public administrations of countries who

have adopted anti-corruption conventions tend to be more affected by

corruption.

A main reason for the under-effectiveness of these laws rests in the fact

that, while being ratified by national parliaments, legislation is not being

implemented in executive agencies most prone to corruption—particularly

traffic police, the police, customs, and tax inspection agencies. Executive

agency-level implementation – in the form implementing regulations – has not

occurred in many Eastern European countries for (at least) two reasons. First,

the development and application of many of the legal principles that help

4. DANIEL KAUFMANN ET AL., WORLD BANK, GOVERNANCE MATTERS: AGGREGATE AND

INDIVIDUAL GOVERNANCE INDICATORS FOR 1996-2005, 4 (2006), available at http://siteresources.worldbank.org/INTWBIGOVANTCOR/Resources/1740479-1150402582357/2661829-1158008871017/gov_matters_5_no_annex.pdf.

5. Id. at 38.

Figure 11: The Distortionary Effects of Anti-Corruption Regulations

civil

servant

colleague

$ tax revenue

draws tax revenue

untied to public goods provision

to create and enforce regulations

other projects

detracts time and resources

from other projects, leading

to real harms

demoralisation

causes decrease in work effort

which can reduce dept efficiency

highly qualified staff exit

distorted markets

sub-optimal service

delivery and excess

taxes create

inefficiencies

regulation

Figure 2: The Distortionary Effects of Anti-Corruption Regulations


insure civil servant accountability (and that prosecutors or instructing judges

rely upon in trial) remains less advanced than in developed Organization for

Economic Co-Operation and Development (OECD) countries. In these under-

developed countries, resort to civil law, administrative law, criminal law, and

contract law is difficult and costly. Unlike in OECD countries, in cases where

the law is silent, no jurisprudential tradition helps inform public administrators

(or the administrative judges who must decide on the legality of their decisions)

to make the decisions that contribute to an effective ―stock‖ of regulation.6

Second, the provisions in these conventions often do not take into account the

political or economic costs involved in implementing the provisions of the anti-

corruption laws these countries ratify. Many of the points—related to the

criminalization of bribery, the cleaning of party finance, and international asset

seizure—prove expensive to implement and run against the interests of

parliamentarians or executive agencies.7

As Eastern European countries begin to draft and adopt implementing

regulations for international anti-corruption conventions, an incentive-based

approach to fighting corruption will produce superior results to an approach

based on ad-hoc regulations because regulations distort both the public and private

sectors.8 In many cases, coordination or cooperation failures between civil

servants create situations where executive regulation can improve productivity

in anti-corruption work.9 However, as shown in Figure 2, in many cases, these

regulations fail to stop the behaviour they target while simultaneously

introducing a wide range of real costs (such as monitoring and enforcement

costs) and economic costs (tied to the opportunity cost of doing something else).

These anti-corruption regulations, which often create a divergence between

actions the civil servant would normally take and the action he or she is

required to take, will detract the civil servant’s attention from other (potentially

more productive) tasks.10 Anti-corruption regulations cost money—money

6. Explaining the existence of legislation and regulatory instruments as the result of costs and benefits represents a fascinating (and under-developed) area of legal scholarship. For more on the costs of bringing various cases to trial, see Simeon Djankov et al., Courts, 118 Q. J. ECON. 453 (2003).

7. At the risk of appearing immodest, Prof. Kasemets and I sketch these arguments in an earlier paper. Bryane Michael & Aare Kasemets, The Role of Incentive Design in Parliamentary Anti-Corruption Programmes, 13 J. LEGIS. STUD. 280 (2007).

8. Many executive agencies in Eastern Europe and the Former Soviet Union have chosen to fight corruption through the adoption of agency-level codes of conduct, conflict of interest rules, asset declaration rules, and broader party finance reform (among the literally hundreds of other executive level regulations ranging from the amount of money civil servants are allowed to carry in their pockets to the number of receipts they must produce for work-related expenses). Passing a regulation represents an instinctive and visceral reaction to a problem in the public sector. For an extremely insightful view of such a phenomenon in a US perspective, see FRANK ANECHIARICO &

JAMES B JACOBS, THE PURSUIT OF ABSOLUTE INTEGRITY: HOW CORRUPTION CONTROL MAKES

GOVERNMENT INEFFECTIVE (1996).

9. I do not address the role of the corruption of the regulatory structures themselves—or the regulator’s role in combating such corruption. For a broader exposé, see Anthony Ogus, Corruption and Regulatory Structures, 26 L. & Pol’y 329 (2004).

10. All regulation is not harmful; as regulations prohibiting civil servants from collecting bribes clearly detracts the civil servant’s interests away from such activities and increases overall productivity. However, regulations which force the civil servant to engage in filling out paperwork or change work location every 6 months (just for the sake of doing so) clearly imposes costs which may be larger than the gains from reducing corruption.


required to talk about them, to write them down and to implement them. These

regulations can often distort markets, not only because they restrict civil

servants’ freedom of movement but also because they impose an extra tax

burden on society. A departmental regulation aimed at reducing corruption must

balance the corruption reducing effects that rule will have with the distortionary effects

on the public management environment and private markets.11

In Eastern Europe, executive agencies are rapidly adopting anti-corruption

regulations that may be more distortionary than socially productive. Figure 3

shows simple statistics summarizing the extent and impact of many anti-

corruption regulations in the Eastern European and Former Soviet region.12 In

several of the Nice accession round countries (Hungary, Poland, Latvia, and

Lithuania), executive agencies have already been writing regulations that adopt

EU decisions or help these agencies comply with acquis obligations.13 Recent

accession round countries (Romania and Bulgaria) have only started to come to

terms with many of the regulatory principles that are common to the older EU

member states.14 Several former Soviet states (Armenia, Georgia, and Russia)

predictably still–despite years of consultants’ advice–have not established a

relatively well-founded regulatory base, owing in part to numerous revisions in

their anti-corruption legislation. In cases where regulations have been adopted

in these former Soviet countries, they often impose more social costs than

benefits–reflecting reliance on the ―stick‖ of administrative punishments

instead of the ―carrot‖ of civil servant incentives.

11. As this is a paper on the legal aspects of anti-corruption regulation, I ignore the obvious observation that regulation must also seek to promote justice.

12. A consistent (and convincing) measure of law enforcement institutions remains sorely lacking in the field of law and economics. For recent attempts, see generally Maria Dakolias, Methods for Monitoring and Evaluating the Rule of Law, in APPLYING THE ―SECTORAL APPROACH‖ TO

THE LEGAL AND JUDICIAL DOMAIN 9 (Ctr. for Int’l Cooperation ed. 2005) 2005, available at http://www.cilc.nll/Conference_publication_2005.pdf.

13. Unfortunately, the most recent comparative analysis of legal work on anti-corruption policy in the region derives from an Open Society Institute study from 2002. OPEN SOCIETY INSTITUTE, MONITORING THE EU ACCESSION PROCESS: CORRUPTION AND ANTI-CORRUPTION POLICY (2002), available at http://www.eumap.org/reports/2002/corruption.

14. In 2007, the Commission found numerous shortcomings with both Romania’s and Bulgaria’s adoption of measures aimed at reducing corruption. Press Release, European Union, Key Findings of the Progress Report on the Cooperation and Verification Mechanism with Bulgaria (June 27, 2007), http://europa.eu/rapid/pressReleasesAction.do?reference=MEMO/07/261&format=HTML&aged=0&language=EN&guiLanguage=en; Press Release, European Union, Key Findings of the Progress Report on the Cooperation and Verification Mechanism with Romania (June 27, 2007), available at http://europa.eu/rapid/pressReleasesAction.do?reference=MEMO/07/262&format=HTML&aged=0&language=EN&guiLanguage=en.


Figure 3: Estimates of Anti-Corruption Regulatory Impacts on Eastern

European and Former Soviet Countries

Country Level ―Stock‖ of

Regulation

(1 to 5)

Level impact

(-5 to 5)

Armenia 1 -1

Bulgaria 3 0

Georgia 1 -2

Hungary 4 3

Latvia 2 1

Lithuania 2 3

Poland 3 2

Romania 2 0

Russian Federation 2 -3

Source: author.

Such data show that right-regulating is clearly more important than under

or over-regulating. More regulation does not produce more effective reductions

in corruption (as Figure 1 showed at an international level). Instead, anti-

corruption regulations should be incentive compatible–creating incentives for

civil servants to follow them. Regulations should allocate liability (or legal

responsibility for damages arising from corruption) to civil servants in such a

way so that civil servants have the incentive to avoid corrupt situations as well

as to report potential corruption in their workplace.

REGULATIONS AIMED AT ALLOCATING LIABILITY FOR CORRUPTION OFFENCES

One area where such right-regulating helps create incentive compatibility

lies in the way that the executive agency allocates liability for non-criminal

cases of corruption. When more than two parties participate in corruption, the

distribution of liability between conspirators and accomplices must be

addressed (as the international conventions mentioned above provide

absolutely no guidance in this area). Corruption often involves only two parties

because incentives are to keep the transaction secret. However, in many

circumstances, bribes are often centralized within the agency, with agents

handing over part or all of the revenue collected to their superiors.15 Corruption

chains exist in the government involving accessories to a crime (those who

actively collect a share of the bribes) and complicity in a criminal offense (those

who know about corrupt colleagues).

In some cases, the superior can be liable for corruption offenses of

15. For a (somewhat outdated) description of the phenomena in Eastern Europe, see Andras Sajo, Corruption, Clientelism, and the Future of the Constitutional State in Eastern Europe, 7 E. EUR. CONST. REV. 37, 39-41 (1998).


subordinates if he or she failed to exercise sufficient oversight. In other words,

if an inspector working at Sofia airport takes a bribe in order to help a naughty

American tourist avoid paying taxes on five cartons of cigarettes, the inspector’s

boss would likely be partly responsible for not keeping a closer eye on the

inspector. Yet, even if the superior exercises reasonable oversight over

employees, he or she may still be prosecuted under the various forms of

respondeat superior existing in various legal codes in many of the countries in

Eastern Europe and the Former Soviet Union. Respondeat superior–an Anglo-

Saxon phrase used to describe a concept that applies to both common law and

civil law legal systems–literally means ―let the boss answer (for the offense).‖

At first glance, Eastern European countries should obviously punish the bosses

of corrupt civil servants (after all, they were not exercising sufficient oversight

over their employees). Yet the strengthening of respondeat superior would have

negative effects on the level of detected corruption. On the one hand, increased

senior responsibility provides senior officials with more incentives to detect and

prevent corruption. On the other hand, criminal sanctions imposed against the

senior official make him or her less likely to investigate vigorously (as he or she

would also face penalties).

These two contradicting factors can be analysed using economics. Figure 4

shows the amount of detection effort and the value of such effort.16 Panel (a)

shows increasing marginal costs of engaging in detection efforts, since the

principal shares criminal liability with the agent. Marginal benefits decrease as

the probability of finding cases of corruption falls—with more and more furtive

glances over the shoulders of the boss’ employees helps prevent potential cases

before they become serious. Given some level of marginal costs and benefits,

superiors will exert an optimal level (or quantity) of detection effort–given at

Q1–at some particular social cost and benefit (described in the Figure as p1).17

Panel (b) depicts the situation–still contrary to the administrative traditions of

most of the countries in the region–where superiors may face less (or no)

liability in cases where they found corruption in their own staff. In such a case,

the overall cost of detection decreases and the level of detection effort increases

from Q1 to Q2 as the marginal costs of detecting corruption fall for all levels of

detection effort.18 A more forgiving policy toward public sector principals can have

16. The economic analysis of anti-corruption work remains controversial (on the grounds that the main function of law remains to promote justice, with only a weak consideration of the cost involved). For readers interested in the economic analysis of corruption, see generally Bruce L. Benson & John Baden, The Political Economy of Governmental Corruption: The Logic of Underground Government, 14 J. LEGAL STUD. 391 (1985). Naturally, Becker remains the canonical reference for this kind of analysis. See generally Gary Becker, Crime and Punishment: An Economic Approach, 76 J. OF

POL. ECON. 169 (1968).

17. For reasons that are too lengthy and tangential to my argument, I will not explain why the social marginal cost of detecting corruption in theory (and possibly even in practice) equals the marginal social benefit. The reader may consult the final chapters of any basic microeconomics textbook for such a rationale. The representation (and quantification) of a level or quantity of detection effort also represents a potential problem for such a simple analysis. Skeptical readers may refer to authors cited below such as Shavell and Polinsky. See infra at note 27.

18. In cases of partial administrative indemnification where the superior would still face some sanction (though softer than currently the case), the superior’s marginal cost curve would also shift out. As a result, further detection effort would also shift out the superior’s marginal benefit curve–making further detection effort more beneficial. As such a situation does not change the ―message‖


Figure 5: A Two Part Complicity

Test for Administrative Corruption

Cases

1. Did the accused accomplice (superiour) likely know about the

corrupt transaction(s) of his or her sub-

ordinates?

2. Did corruption by the accomplices’

sub-ordinates result in significant harms to individuals not party to the

corrupt transaction (or expose such

individuals to such harms)?

the counter-intuitive effect of decreasing corruption.

Thus, executive agencies in many of the countries in the region should

establish a regulatory list test establishing when a superior (or colleague) may–

to the social profit–escape the punishment inherent in his or her administrative

relationship with a corrupt

official. The international

conventions take an unclear

position on the exact extent to

which the failure to monitor

subordinates (or even other

colleagues) constitutes punishable

complicity. As shown in Figure 4,

provisions providing for an

escape from the liability inherent

in a principal-agent (or duty of

care) relationship may result in

more anti-corruption detection

effort by civil servants. Yet civil

servants of any rank should not

obtain blanket immunity from prosecution (as Figure 4 also shows that passing

through some liability may increase the superior’s marginal benefits of exerting

detection effort).

A legal test should be in place at the regulatory level to determine whether

an individual’s detection effort qualifies for exclusion from a disciplinary or

administrative prosecution.19 Figure 5 shows the two basic ingredients for such

of Figure 4, I leave out these details to keep the argument simple.

19. Such a test would only apply to disciplinary or administrative prosecutions, as criminal cases follow already well-defined rules of procedure.

quantity of

detection effort

Figure 4: Optimal Public Sector Principals’ Effort in Detection Corruption Offences

price/value

of detecting

corruption

marginal

costs

(related to

shared

criminal liability)

marginal

benefits

(in reduced

likelihood

of corruption)

quantity of

detection effort

price/value

of detecting

corruption

marginal costs’

marginal benefits

Q1

p1

p2

Q2


a regulatory test. The first part of the test establishes whether the superior

exerted sufficient effort in detecting corruption among his or her subordinates

(or peers). If the superior likely knew about the corruption of his or her

subordinates, then the person clearly either served as a direct accomplice or

engaged in an inefficient level of detection effort (thus being guilty of

negligence).20 The second ingredient in the test–a typical harms test–basically

establishes jurisdiction over the person’s case. If the accomplice (the superior)

proved negligent (as exercising insufficient detection effort) and that negligence

resulted in serious harm to others or the agency with which he or she works,

then criminal or civil liability ensues.21 Regulatory jurisdiction (and thus the

test contained in Figure 5) would clearly be unavailable to members of the

agency’s disciplinary committee if significant harm ensued from the superior’s

lack of detection effort.

Regulatory jurisdiction–and the jurisdiction of corruption cases in general–

serves as one of the key drivers of the marginal costs and benefits of anti-

corruption work in Eastern Europe. As shown in the first section of this paper,

the international conventions–with their stress on the criminalization of

corruption–have proved ineffective in fighting corruption. In the second

section, I argued that anti-corruption law–regulations in particular, but

legislation as well–should result in excess social benefits (compared with the

costs of writing and enforcing anti-corruption law). In this third section, I

tackled a specific legal concept (complicity and the responsibility of a civil

servant’s superiors) in order to show how regulation that takes into

consideration regulatory costs and benefits can lead to more effective anti-

corruption work in Eastern Europe. Throughout the article, I have discussed

the non-criminal investigation and prosecution of corruption (and the

regulations that may support such a process). What determines whether a

corruption case should be classified as a criminal–rather than administrative or

disciplinary–offense?

AGAINST CRIMINALIZATION? JURISDICTION FOR ANTI-CORRUPTION OFFENCES

The international conventions against corruption as embodied in the United

Nations, Council of Europe and Organization for Economic Cooperation and

Development (―OECD‖) have increasingly pushed for the criminalization of

20. The word ―likely‖ in the test poses difficulties for any disciplinary committee. In practice, the committee might use a two step approach to establish likely guilt. First, the committee would establish the optimal level of detection effort (naturally in approximate and qualitative terms) provided by the incentives the superior faced. Second, if the superior had incentives to engage in a relatively high level of detection effort, then the disciplinary committee should establish guilt based on what the Anglo-Saxons refer to as the ―balance of probabilities.‖ Under a balance of probabilities test, if the accused accomplice had a 51% probability (as judged by the disciplinary committee) of knowing about the corruption of his or her subordinates, then the accused would be found guilty of negligence.

21. As in cases of theft or other offenses, a particular level of harm often determines whether an offense is statutorily defined as criminal or civil. The most obvious measure of harm in cases involving bribery consists of the financial value of the bribe involved. In the early stages of using this test, judges and administrators may interpret the phrase ―significantly adversely‖ and, over time, embody their deliberations in the anti-corruption law as a threshold level.


bribery and corruption. Such a push for criminalization stems from the

common practice among developed OECD countries with effective judicial

systems to treat corruption as a crime. However, in countries with weak

judicial systems, the criminalization of corruption offenses (as imposed by these

international conventions) poses serious problems for successfully prosecuting

corruption offences.22 Criminal cases require extensive (and expensive)

investigation by the police and/or prosecutor’s office (depending on the

country). Yet, these government services severely lack budgetary resources–

and high levels of corruption in their judicial system further reduces the

likelihood of successful prosecutions of corruption in other sectors. Moreover,

(as previously discussed) as superiors are legally and administratively

responsible for the corrupt activities of their subordinates, high-level officials

have been unwilling to strenuously investigate complaints about corruption–in

order to avoid prosecution themselves!

Two other arguments militate against the blanket criminalization of

corruption in Eastern Europe and the Former Soviet Union. First, the amount of

proof required for a successful criminal conviction reduces the range of cases

that criminal investigators pursue. Many times, investigators decline cases

because they do not believe that they will be able to collect enough evidence to

convince a court beyond a reasonable doubt about the suspect’s guilt.

Administrative sanctions against corruption based on a balance of probabilities

standard for successful conviction can, in certain circumstances, provide greater

deterrence against corruption than the ostensibly stronger criminal standard,

which requires proof beyond a reasonable doubt.23 Second, by increasing the

penalties for bribery, criminalization can increase the equilibrium bribe

payment. Obviously, bribees need to ―price in‖ the additional costs they face if

they should be detected (in serving prison time). Less obviously, once a bribe

has been offered, the civil servant bribee can blackmail the briber up to the

point where the bribe paid equals the money value of criminal sanctions against

the briber. In many cases, a suspect will pay any amount to avoid serving a jail

sentence in a Tbilisi, Vilnius, or Budapest prison.24

The difficulty of prosecuting corruption under a criminal burden of proof–

as well as the large number of mitigating circumstances (which corrupt civil

servant defendants can use as reasonable and reliable defenses)–suggests that

other jurisdictions may provide higher benefit and lower cost venues for

adjudicating corruption related cases.25 Figure 6 shows three levels of potential

22. Some authors paint a rosier picture of the role of the anti-corruption conventions in changing legal approaches to fighting corruption. See Steven Salbu, A Delicate Balance: Legislation, Institutional Change, and Transnational Bribery, 33 CORNELL INT’L L.J. 657, 660 (2000).

23. For years (and particularly inspired by the UK’s treatment of corruption), legal scholars have considered the relaxation or reversal of the burden of proof in corruption cases. See generally Ndiva Kofele-Kale, Presumed Guilty: Balancing Competing Rights and Interests in Combating Economic Crimes, 40 INT’L LAW. 909 (2006).

24. While this paper focuses on the economics of anti-corruption, the economics of corruption itself has blossomed into a wide and interesting field. For a classical reference, see generally SUSAN

ROSE-ACKERMAN, CORRUPTION AND GOVERNMENT: CAUSES, CONSEQUENCES AND REFORM (Cambridge University Press 1999).

25. Very few legal scholars have focused on secondary or delegated legislation aimed at fighting corruption. For one notable exception, see Benjamin B. Wagner & Leslie Gielow Jacobs,


―seriousness‖ (or responsibility) existing in the law of each country in the

region (albeit in a highly disarticulated form, which appears in different parts

of each country’s law). As shown in Figure 6, each level of jurisdiction

(―responsibility‖ in simpler terms) corresponds roughly with the harms (or

potential harms) attendant with each type of corruption-related offense. In

most of the countries in the region, various executive agencies (such as customs,

police, or others) may investigate or deal internally with between five and

twenty cases of corruption per month (naturally Russia as a large country will

deal with more cases than Latvia). Whereas before, internal investigators dealt

with cases themselves (for better or worse), legislative changes in all countries

in the region have resulted in these agencies ―passing over‖ case files to the

criminal investigators. Such legally mandated ―passing the buck‖ may explain–

in part–why the international conventions are failing to help reduce corruption

in Eastern Europe.

Figure 6: The Harms and Standard of Evidence Required for Corruption

Remedies

Evidence

requirement

Types of

remedies

Advantages

Managerial

(defined in

Agency’s Law

or Code)

Suspicion of

corruption.

Written warning,

reassignment,

pass to

administrative or

criminal levels.

Failure to prosecute

opens to

administrative and

criminal liability.

Administrative

(defined in

Admin Law)

Balance of

probabilities.

Warning,

reassignment,

fine, firing.

Failure to detect large

numbers gives

administrative court

liability, contributory

factors/ accessory to a

crime.

Civil

(defined in

Civil Code)

Depends on

legal tradition.

Payment of

damages and

compensation.

Addresses economic

harms, provides for

deterrence that jail

time does not.

Criminal

(defined in

Criminal Code)

Beyond a

reasonable

doubt.

Fine, censure,

prison.

If criminal prosecution

does not succeed

(because of corruption

in judiciary), then rely

on two other levels to

obtain partial

prosecution.

Note: For each Code, please see www.legislationline.org.

Criminal investigators and prosecutors should only be involved in corruption cases

Retooling Law Enforcement to Investigate and Prosecute Entrenched Corruption: Key Criminal Procedure Reforms for Indonesia and Other Nations, 30 U. PA. J. INT’L L. 183 (2009).

http://www.legislationline.org/


in Eastern Europe when the benefits to society of treating the case as a crime exceed the

costs.26 First, petty corruption cases should not receive the same time and

resources as grand corruption. Governments, like individuals, face budget

constraints and when their expenditures exceed their income, they go bankrupt.

Second, in a system where over-regulation causes incentives for the majority of

the population to be corrupt, the criminalization of everyone is tantamount to

the creation of a nation of outlaws. Third, ―devolving‖ investigatory and

prosecutorial powers to executive agencies–particularly law enforcement

agencies with already existing legal powers in these areas–can both create anti-

corruption competencies throughout the government as well as reduce the cost

of anti-corruption work, since an investigator in the Internal Affairs

Department of Customs is likely to earn less than the Chief Inspector in the

Ministry of Interior.

A proper anti-corruption regulation clearly assigns jurisdiction over

corruption cases to the least-cost, highest-benefit jurisdiction. Managerial

responsibility aims at tackling incentives leading to corruption and small

corruption offenses. Managerial jurisdiction results in cheap investigations,

which require little formality. Administrative responsibility applies a civil law

burden of proof (allowing executive agency’s services to deal with high risk

areas of corruption where obtaining proof is difficult or expensive).

Administrative cases can be processed quickly and the relatively light penalties

make bribing administrative judges generally unprofitable. Criminal

responsibility applies all the standard procedures as envisioned in the criminal

code (and to the extent applicable the anti-corruption law).

Yet, because of differences between the social and agency-level harms involved in

corruption cases, strong incentives exist to misallocate jurisdiction over corruption

cases. Figure 7 shows the logic behind the misallocation of cases (a

phenomenon that most actual investigators and prosecutors in Eastern Europe

and the Former Soviet Union complain about often). The heavy upward sloping

line in Figure 7 represents the agency’s inherent or ―technical‖ marginal costs of

prosecuting corruption cases. Bigger corruption cases–namely those cases

involving more money or more economic harm–require more investigatory and

prosecutorial resources. Huge cases like the Siemens corruption scandal require

more resources than cases involving small €5 notes slipped to a doctor. To start

even the most basic case requires some investment of resources, labeled as V0.

26. Many critics would argue that justice has no price. However, these critics often live in societies whose justice systems are well-financed. In middle and lower-income countries, unfortunately, economic choices must be made about the ways in which scare resources are used. The critics’ refusal to address the economics of an issue related to justice has resulted in long queues on court dockets, long waiting times in filthy jails, and a court system concentrating opportunities for bribe payments around criminal prosecutors and judges.


Corruption comprises an economic activity involving negative economic

externalities. Namely, the social harms arising from corruption exceed the

private harms to individuals or the Agency affected. The Customs Agency may

lose €10 million in collected revenue from under-valuation due to the payment

of bribes, but the economic distortions those bribe cause on other traders (more

efficient traders losing contracts, becoming unable to obtain imported inputs

and so forth) can easily exceed €100 million. Figure 7 shows the gap between

these social and private (namely Agency-level) harms by two upward curving

marginal cost curves. The Agency’s managers will want to ―take care of‖ low

value anti-corruption cases themselves because they see that their capacities of

investigation and prosecution equate with the level of harm to the Agency (as

shown by point C1 in Figure 7). At point C1 though, the social harms present

much greater losses to society than agency’s managers internalize. At the other

extreme, large corruption cases can be tackled by an agency-wide special task

force. As represented at point C2, the agency has probably lost a large amount

of revenue due to a particular case of corruption and seeks either restitution or

at least public recognition for prosecution corruption among its own ranks. Yet,

again, the corruption case generating harms represented by C2 in Figure 7

represents a case where the agency should turn the case over to criminal

investigators and prosecutors.

Anti-corruption regulation should provide for agency-level jurisdiction

over corruption cases only when their marginal costs equal the social marginal

benefit of prosecuting these cases at the agency-level instead of at the criminal

level. As shown in Figure 7, an administrative level prosecution–involving the

Agency’s disciplinary committee or an administrative tribunal hearing–should

ensue when the cost to the administrative instance of hearing the case equals

the harm the case poses to society (and not just the Agency itself). In the case

shown in Figure 7, despite any provisions in the law, managers will have incentives to

Figure 7: The Optimal Division of Corruption Remedies Between Jurisdictions

social

marginal

costs

Agency’s

(private)

marginal

costs

detection and

prosecution

marginal costs

departmental

remedies

administrative

remedies

criminal

remedies

corruption harms

marginal

costs

C1

Optimal level of

prosecution

Quick and

easy prosecution

Desire to keep it

“in the house”

C* C2

Vp

Vs

V0


claim jurisdiction over cases which they should refer to the Internal Affairs Department

(or the relevant Agency-wide disciplinary committee) as well as claim jurisdiction over

cases which should be turned over to the police.

Without the proper form of anti-corruption finance, agencies will not have

the incentives to assign corruption cases to their optimal jurisdiction.27 As

shown in Figure 7, the disparity between social and agency-level costs and

benefits will prevent agencies from working in the overall interest of the

country. Individuals responsible for fighting corruption in an executive must be

provided with incentives in order to exert sufficient effort in investigating

corruption and/or turning over the case the highest-benefit, lowest-cost

jurisdiction. The social benefits will accrue as additional resources–either in the

form of increased taxes or as the result of confiscated proceeds from corruption.

Clearly, these resources should be allocated in order to encourage (or provide

incentives) for anti-corruption work, leading to a discussion of anti-corruption

finance.

ANTI-CORRUPTION FINANCE: CREATING INCENTIVES TO EXPOSE CORRUPTION

The social benefits of fighting corruption–which translate into financial

benefits like more taxes and the confiscation of the proceeds from corruption–

should be used to encourage further anti-corruption work. At present, anti-

corruption programs in all the countries in the region are grossly under-

financed. Latvia, Lithuania, Hungary and Poland have received large amounts

of (unsustainable) anti-corruption donor assistance. Romania and Bulgaria

have received less (though partly because of the large amounts of other finance

aimed at accession, which would have had indirect impacts on corruption).

Georgia, Armenia and the Russian Federation have funding for a wide-variety

of ad hoc donor supported programs. In all these countries, actual budgetary

allocations of anti-corruption work are usually determined on an historical

basis (looking at last year’s budget and adding a bit). Yet none of these

approaches follows an incentive based approach, leaving corruption fighters

with few incentives to go chasing bribe takers. Promising anti-corruption

fighters the same amount of money every year (or a bit more) does not motivate

them to work harder.

27. Authors writing on issues related to the optimal level of law enforcement frequently omit any discussion of incentives and finance. See generally A. Mitchell Polinsky & Steven Shavell, Corruption and Optimal Law Enforcement, 81 J. OF PUB. ECON. 1 (2001).


Theory suggests that performance-based budgeting provides the strongest

incentives for reducing corruption. Figure 8a shows a number of anti-

corruption budgetary finance rules. The figure shows the structural (or

naturally inherent) relationship between the level of corruption and the

corresponding anti-corruption effort as a heavy downward slopping line and

anti-corruption effort

level of corruption

fixed finance case

political

mandate case

Figure 8a: The Case for Performance-Based

Anti-Corruption Budgets

structural trade-off

between corruption and

anti-corruption

ACff

Cpm

anti-corruption

effort

level of corruption

fixed finance case

political

mandate case

Figure 8b: The Case for Performance-Based

Anti-Corruption Budgets

structural trade-off

between corruption and

anti-corruption

performance-based

budgeting

C*

AC*


three different anti-corruption financing rules.28 In the case of a fixed budget to

a department (such as the Internal Affairs department of a Ministry of Interior),

the anti-corruption level of effort is fixed by the department’s line-item budget

for anti-corruption work. Such funding, depicted as a vertical line, establishes a

fixed level of anti-corruption work–regardless of the level of corruption.

Georgia, Armenia and the Russian Federation represent countries where such

an approach remains dominant. In the case where funding is given based on

political events (a general outcry against corruption) or based on the

administratively perceived need for anti-corruption funding, funding is tied to

a particular level of corruption that may not correspond to the amount of actual

anti-corruption work required in the department. Such funding–depicted as a

horizontal line and labeled as the ―political mandate case‖–often only occurs at

high levels of corruption (as governments tend not to fund anti-corruption

where no apparent need exists). Lithuania represented an early example of

such an approach (though recently, anti-corruption finance has become more

like the fixed-finance case). Poland, Hungary and Latvia have, depending on

the point in time, have all seen anti-corruption budgetary allocations wax and

wane with popular perceptions of corruption (with relatively little impact on

reducing actual corruption).

Tying anti-corruption budgets to performance helps allocate resources to

agencies which can best use them.29 Figure 8b shows an example of such a rule

and the optimal level of anti-corruption effort for an agency working on

fighting corruption. A performance-based finance rule would provide funds to

agencies to the extent that they demonstrated that they have reduced

corruption (thus the negative slope of the line). Clearly, when anti-corruption

investigators or other law enforcement officials are able to reduce more

corruption (per unit of effort) than the resources they require as part of their

salary, they should receive more funding (as represented by a steeper sloping

structural trade-off line as compared with the money given by the performance-

budget rule). When these corruption busters receive more compensation (per

unit of effort), then the value they generate in corruption busted, then they

should receive marginally less funding. Clearly, the intersection of each

financing rule with the structural trade-off determines the equilibrium level of

corruption and anti-corruption. Under the assumptions of the simple model

presented in Figure 8, a performance-based budgeting rule will result in the

highest amount of long-run, sustainable anti-corruption effort.30 In this case,

28. The downward sloping heavy set line represents a structural relationship (or overall technical efficiency) gauging the extent to which cops on streets, public auditors slouched over desks, and others help dissuade people from engaging in corruption or help prosecute them. For the purposes of this article, such ―technical efficiency‖ represents know-how, skills, and other exogenously given non-incentive based attributes.

29. Increasing evidence suggests that law enforcement officials respond to high powered incentives they are given. See Luciana Echazu & Nuno Garoupa, Corruption and the Distortion of Law Enforcement Effort, AM. L. AND ECON. REV. ADVANCE ACCESS, January 4, 2010, 17-18, available at http://aler.oxfordjournals.org/cgi/reprint/ahp020v1.

30. While a discussion of comparative statics involving Figure 8 extends beyond the bounds of this paper, the reader can see that the slope of the financial rule between corruption and anti-corruption must be greater than the slope of the structural trade-off between these two variables (in order to guarantee continued investment in anti-corruption).


over time, funding is expected to converge to the level corresponding with the

structural trade-off. In practice, in all the countries in Central Europe and the

Former Soviet Union, governments follow the opposite logic—restricting funding as

anti-corruption work becomes more successful (on the grounds the problem is less

pressing relative to other problems). Such funding rules provide anti-corruption

departments with incentives to avoid fighting corruption (and in theory to contribute to

corruption).31

Qui tam rewards serve as another example where anti-corruption regulation

may lead to strong incentives to fight corruption.32 Qui tam legal provisions,

which allow any individual to sue the government on behalf of the interests of

the government, serve as a bounty for corrupt officials and businessmen.33 The

reward for individuals reporting cases of corruption would be related to the

benefit to society (known as the social benefit) that such qui tam denouncement

produce. For example, suppose a police officer offers to ―ignore‖ a driving

offence commitment by a middle-aged Russian woman driver on the streets of

Moscow in exchange for €40 (the offense could be real or a false rent-seeking

accusation). How much money should the Russian state pay the woman to

denounce the police officer (called GAI-ishnik in Russian because the letters for

the traffic police are GAI)? Clearly the reward needs to be sufficient enough for

the woman to risk her safety (in fear of potential retaliation by the GAI-ishnik).

Figure 9 shows that the woman can be offered up to a sum (€23,540), which at

first glance seems exorbitant given the miniscule €40 bribe requested by the

GAI-ishnik. While the numbers in the example portrayed in Figure 9 are

examples (used only for the sake of illustration), they demonstrate a simple idea

from economic theory–that the social benefits of fighting corruption

significantly exceed private (or individual) costs to contributing to the fight

against corruption. If individuals could share part of that social benefit (as

personal compensation), then strong incentives would produce a much larger

volume of investigations and prosecutions in the countries in the region.

31. In many OECD member countries (particularly the ones with relatively low levels of corruption), government departments–through budgetary bargaining, outsourcing and other methods–have some tie (albeit relatively weak) between the contribution their anti-corruption efforts make to the state budget and their budgetary allocation for the year. Such a mechanism represents an indirect ―bounty‖ system such as the one discussed more directly in the text.

32. I treat the issue of qui tam as a regulatory issue rather than as a legislative issue for two reasons. First, most EU member states politically find qui tam unpalatable–excluding the possibility of qui tam friendly legislation in the near future. As such, a regulatory ―surrogate‖ (relying on legislative provisions that would offer regulators room for maneuvering) would be the best present strategy for adopting qui tam-like instruments in the EU. Second, in many EU member states, a set of rewards and benefits can be discretionarily conferred at the executive agency level which would replicate a qui tam reward.

33. The actual effect of these bounties will depend on the extent to which corrupt officials and businessmen can ―price in‖ the bounty and the increased probability of being caught during the bribe negotiations. If they can pass through these costs, then qui tam provisions may only serve to further redistribute income instead of lead to a reduction in the incidence of bribery. For a game theoretic discussion of the design of such a system, see Robert Cooter & Nuno Garoupa, The Virtuous Circle of Distrust: A Mechanism to Deter Bribes And Other Cooperative Crimes 13 (2 Berk. L. & Econ. Working Papers, 2000), available at http://escholarship.org/uc/item/83c0k3we.


How much money should the government of Lithuania, Hungary, or

Russia (among others) pay denouncers of corruption as qui tam rewards? Figure

10 illustrates the logic more fully.34 Panel 10(a) shows–on a simple line–the

costs and benefits involved in a simple bribery case in which a third-party

denounces a corrupt official in exchange for a reward. The panel shows the

bribe level (on a line such that positions to the right represent higher bribe

amounts) and the corresponding amount of social harm. For example, a person

bribing a customs officer may pay €100, which causes €600 in harm to other

businesspeople (who lose sales). Panel (b) shows a reward level that is higher

than the whistle-blower’s costs (such a reward must always be higher,

otherwise the whistle-blower has no incentive to denounce corruption). For

example, of the €100 bribe, the denouncer may receive €80 and cost the

denouncer €40 in time and effort to find and report this incidence of bribery. At

the level of rewards and costs portrayed in panel (b), the loss to society of

corruption is represented only by the denouncer’s costs (as the other harms are

recoupable). The denouncer’s profit is represented by the difference between

the reward level and the denouncer’s costs. The government compensates the

victims directly harmed by the corrupt act–representing a gain to these victims

(as society now avoids this social harm and possesses the extra resources in

order to make these compensations). Parties who are indirectly harmed will

benefit from the discontinuation of corruption (as shown by the area shown in

34. The use of qui tam legal provisions to provide high-powered incentives to fight corruption remain one of the most exciting research areas of anti-corruption policy. See generally Aaron Petty, Note, How Qui Tam Actions Could Fight Public Corruption, 39 U. MICH. J.L. REFORM 851 (2006).

Figure 9: Calculating the Social Gain of a Qui Tam Accusation by

a Muscovite Woman against a GAI-chik

Woman

Driver

Gai-chik

€ 4,100Other drivers see effect and engage

in trade creation due to lower driving

“transactions costs” € 3,500

Lady less afraid to transact

with others (like go shopping)

€ 600

Trade Creation Effect

Gaichiki’s friends see they may need

to pay – so take less bribes € 7,000

Other gaichiki rely less on

generalised corruption to provide

impunity € 5,000

Possible damages from gaichik (or

evidence of other harms) € 400

Gaichik’s Friends

€ 9,000Gaichik reduces level of bribe

taking - € 2,000

Dissuasion Effect

€ 10,000Gaichik engages in real

policing like fining criminal

speeders € 5,000

Forbearance Effect

€ 23,540Total Social Benefit

€ 440Lady recovers € 40 Revenue Loss Effect

TotalsGaichikEffect

Other

Driver

Other

Driver

Other

Traffic

Police

Other

Traffic

Police

Other

Driver Other

Traffic

Police


Panel 8(b) as ―gain to third parties‖) as €500 in this particular example.

In cases where the reward requested by the denouncer exceeds the amount

of money involved in the bribe transaction, some victims who are indirectly

harmed will have incentives to make a Pareto-improving payment to the

denouncer.35 In panel (c), the denouncer’s costs are less than the bribe paid,

though the reward is much higher. Thus, the denouncer may incur €80 in costs

(for example) and receive a reward of €200 (which is less than the €600 in

overall harm to all producers). As a result, part of the denouncer’s bounty will

be paid by individuals directly harmed by the bribe and partly by individuals

indirectly harmed. However, because the reward is less than the overall gain

accruing to third-parties, some of these third-parties will benefit.36 Panel (d)

depicts the situation where the denouncer’s costs are higher than the amount

involved in the bribe transaction. For example, the denouncer may incur €200

in costs and receive €400 in rewards, though only €300 in bribes was paid. In

this case, as long as the reward paid is less than the value of the overall harm to

society, third-parties will benefit by contributing to the denouncer’s bounty. In

this case, third-parties (or society as a whole) compensate the denouncer and

third-parties gain less than when smaller rewards are paid.37

35. In theory, individuals who are indirectly harmed will make Pareto-improving payments to the denouncer. Usually in practice, however, these third-parties are completely disassociated with the corruption case (and often unable to understand the harms that they experience). Thus, compensation to the denouncer based on overall social gain comes from the government budget. In theory, third-parties who are most harmed by bribery would be those parties most willing to pay compensation to the denouncer. However, even in theory, the policymaker will, only with extreme difficulty, predict which third-parties will benefit most from the discontinuation of corruption (and ask for contributions from these specific individuals).

36. In theory, in the presence of low transaction costs, third-parties who are harmed by corruption can form a coalition or class-action in order to divide the gains relatively equally amongst themselves.

37. As an interesting aside, in the case where no detection or prevention is attempted, the bribe represents a simple redistribution of resources between briber and bribed (thus no economic loss occurs, except for the harms to third-parties). In the case where a bounty-hunter attempts to collect

bribe level social harm

denouncer’s costs

€40


reward level

€ 80

profit to denouncer

from bribe proceeds

€40

extra money returned to

government/owner -- €20

gain to third-parties

loss to society

Figure 10: The Optimal Theoretical Qui Tam Award in Corruption Cases

(a)

(b)

€600

€600

€100


In cases where the reward is set too high, it can distort incentives, causing

qui tam provisions to do more economic harm than good. Panel (e) shows the

case where the reward paid to the denouncer is higher than the direct or

indirect harms of bribery. In this case, the government must use resources from

other activities in order to compensate the denouncer, resulting potentially in

lower social returns to other activities. Because the denouncer will have an

opportunity cost (such as working in a company), the reward could distort the

individual’s incentives, resulting in this individual moving away from other

productive activity.

Effective anti-corruption regulations provide the same kinds of incentives

to civil servants to denounce corruption (and to increase work effort aimed at

investigating corruption). However, paying civil servants to do their duty

potentially creates negative incentives within the public sector because of the

a reward, his or her costs represent a efficiency-decreasing transaction cost.



denouncer’s costs

€80

reward level

€ 200

denouncer’s costs

€ 200

reward level

€ 400

profit to denouncer from

bribe proceeds

€120

third-party subsidy to denouncer

Figure 10 continued: The Optimal Theoretical Qui Tam Award in Corruption Cases

(c)

(d)

gain to third-parties

€ 400

loss to society third-party subsidy to denouncer

gain to third-

parties

€ 200

bribe level

€ 300


reward level

€1,000

next-best

activity

Figure 10 continued: The Optimal Theoretical Qui Tam Award in Corruption Cases

(e)

diversion of public resources

from other uses

diversion of denouncer

from other productive

activities

indirect and indirect benefits


two negative effects describes in panel (e), namely distorting the civil servant’s

time and government resources. As discussed above though, it is possible to

design a compensation package which provides blunted financial incentives (in

the form of promotion or perquisites) without diminishing incentives to comply

with work obligations. Thus, the optimal high-powered incentive scheme aimed at

encouraging civil servants to actively fight corruption blunts the incentives paid to civil

servants to the point where the temptation to divert one’s time and government

resources into the investigation and prosecution of corruption offences that have low

expected return are minimized.38

PUTTING IT DOWN ON PAPER: DRAFTING AGENCY ANTI-CORRUPTION

REGULATIONS FOR CENTRAL AND EASTERN EUROPEAN COUNTRIES

All of the issues presented in this paper can be brought together in an

Eastern European executive agency’s regulations that implement the national

anti-corruption law (and thus the CoE, UN and OECD international

conventions against corruption). National legislation provides general

principles and guidelines that the police, customs, minister of health and other

executive agencies must implement. For example, according to legislation in

place in all the countries in the region, executive agencies must provide mutual

assistance to other countries in international corruption cases. However,

national legislation purposely leaves a large amount of regulatory discretion to

executive agencies to implement each article of anti-corruption law as they see

fit.39

Anti-corruption regulations–like the legislation that provides the over-

arching principles for such regulation writing–should strike a balance between

defining more clearly the operating methods of implementing legislation and

the constraints those regulations pose on an agency’s staff in taking bribes (or

engaging in corruption). Figure 11a shows the way in which these two effects

determine the level of anti-corruption regulations in a country (taking a

relatively low corruption OECD member state as an example). In the figure, as

executive agencies pass more and more pages of anti-corruption regulations,

the cost per page falls (the as the main issues can be tackled in a few pages).

Additional pages (or separate instructions) serve to ―tighten the screws‖

around civil servants–constraining them less and less and reducing less

corruption on the margin. Figure 11a depicts these downward sloping marginal

38. The optimal payment, which will be less than the equilibrium payment that equals the private plus social costs imposed by a particular corrupt transaction, will clearly depend on the civil servant’s income elasticity of labor supply and the extent to which the civil servant internalizes the externality through a public service motivation or altruism. For empirical estimates of such a public service motivation, see Philip E. Crewson, A Comparative Analysis of Public and Private Sector Entrant Quailty, 39 AM. J. OF POL. SCI. 628, 636 (1995). Such a payment will necessarily reduce the resources available to compensate victims of corruption crimes and other corruption offenses.

39. I use the term ―anti-corruption law‖ because the various pieces of legislation governing the ways that a country fights against corruption vary enormously across countries. Countries such as Latvia and Lithuania, for example, rely mainly on a specific anti-corruption piece (as a single piece of legislation). Other countries, such as Russia (until recently) have relied on various legal provisions scattered across the Criminal Code, Administrative Code, Public Procurement Law, and various other pieces of legislation.


costs as MC1. On the other hand, more pages of anti-corruption regulations add

precision, outline complicated procedures for tackling corruption and can even

serve as a useful surrogate for legislation (when such legislation is extremely

vague). Figure 11a shows these upward sloping marginal benefits as the curve

MB1. Naturally, an executive agency will write anti-corruption regulations

until the cost per page (or instruction) of writing these regulations equals the

benefit in fighting corruption. In other words, the marginal costs of regulation

equal the marginal benefits at an optimal ―quantity‖ (as measured in pages,

number of instructions or some proxy measuring the quality of regulation

writing) depicted in the figure as QOECD.

For OECD member countries (countries with already relatively low levels

of corruption), the effects of passing more anti-corruption laws only causes

agencies to rewrite their already existing regulations. The marginal benefits of

their present regulations decrease (as the guidance offered by legislation

replaces the guidance to officers previously offered by Agency-level

regulation). Figure 11a shows such an effect as a shifting in of the marginal cost

curve to MB2. However, the costs to the Agency of their regulations on the

books increases as new regulations come into force that replace the ―tired and

true‖ ways these agencies developed over decades to fight particular types of

corruption in their ranks. While the exact effect on the ―stock‖ (or ―quantity‖)

of regulations will depend on particular details, the level of regulations should

stay about the same. In my own reviews of regulations from police departments

in the USA, France and the UK, after the wave of legislative changes took place

in the early part of the 2000s, the predictions of Figure 11a seem relatively

sound.

For countries in Eastern Europe and the Former Soviet Union (such as

Georgia, Armenia, Hungary and so forth), the quantity also remains the same–

because of Agency budget constraints. Figure 11b shows the quantity of anti-

corruption regulations and the costs and benefits of adjusting to legislative

“quantity” of anti-corruption regulation

Figure 11a: Passing a New Law: Effects on Anti-Corruption Regulations

for an OECD Country

price/ value

of AC

regulation

Qoecd

marginal costs

constraining effect

marginal benefits

guiding effect

MC1

MC2

MB1

MB2


changes in these Eastern European and Former Soviet countries. Anti-

corruption legislation in the countries in the region (with the exception of

perhaps Romania) is more abstract and less well-defined than in other

countries. These laws tend to be shorter, the wording more diffuse and the

principles (while often the same), stated in more ambiguous language. As such,

the marginal benefit of anti-corruption regulations should increase as new anti-

corruption legislation passes the Duma (Russia), Seimas (Lithuania),

Országgyűlés (Hungary) and so forth. Figure 11b shows the optimal level of

anti-corruption regulation in the region at QEE/FSU–well above the optimal level

in OECD member countries. Legal departments should have the incentives to

work vigorously on new and more specific anti-corruption regulations.

However, because of rigidities in legal institutions in many of the countries in the

region, the quantity of anti-corruption regulations remains well below the optimal

level–or even completely unchanged. In theory, the marginal costs of engaging in

anti-corruption regulation should fall because of ―better‖ legislation. However,

in practice, most Eastern European states (and particular those in the Former

Soviet Union) leave regulations in place. Their legal departments function

poorly (if at all) and staff working in these legal departments have no incentives

to draft revised regulations (given broader problems with the public sector pay

and promotion framework). As such, marginal costs should–hypothetically–

shift inward (as shown by the MChypothetical line). In practice, though, the

equilibrium level of regulation, in the absence of resource constraints and

responding only to incentives, would be at Q*. However, these countries have,

in practice, a much lower ―quantity‖ of anti-corruption regulations. A visit to a

customs office in Tbilisi, Riga, or St. Petersburg will reveal regulations

consisting of only a few pages and written in very informal language.

Naturally, the necessary regulations remain poorly written (or unwritten)

because legal staff (what few work in the agency) are not paid enough. In other

words, resource constraints keep the level of anti-corruption regulations at

Qconstrain.

“quantity” of anti-corruption regulation

Figure 11b: Passing a New Law: Effects on Anti-Corruption Regulations

for an Eastern European Agency

price/ value

of AC

regulation

Qoecd

marginal costs

constraining effect

marginal benefits

guiding effect

Message: Laws even more abstract...needs even more regulations

but low budgets, so crappy regulations...

Qconstrain

MC1

MB1

QEE/FSUQ*

MChypothetical

MB2


The suggestions made in this paper should, in part, help the countries in

the region raise their anti-corruption regulations toward the optimal level. In

the first part of the paper, I looked at the types of ―value-subtracting‖

regulations and proposed the adoption primarily of welfare-improving

regulations. The second section looked at ways of improving compliance with

anti-corruption regulations through designing incentive compatible regulations

(particularly those that allocate legal liability for corruption offenses at the non-

criminal level). The third section looked at cases where the agency specifically

should adopt the regulations described in Figure 11. The agency should,

following the thesis provided in this paper, be assigned jurisdiction over a non-

criminal corruption case when the social benefits outweigh the costs. The

fourth section noted that providing agency-level anti-corruption fighters with

the incentives to investigate and prosecute cases requires finance tied to

performance. Namely, an agency’s budget for fighting corruption (whether in

the form of salaries to members of an Internal Affairs unit or simply for printing

posters) should be tied to the spending agency’s effectiveness at fighting

corruption. Rewards can incentivize civil servants as well as private citizens–as

seen in the case of qui tam rewards. Through incentive-based mechanisms, the

countries in the Eastern European and Former Soviet region can improve the

regulatory quality (and quantity) of their anti-corruption instructions, orders,

and decrees.

CONCLUSIONS

Through incentive-based mechanisms, many countries in Eastern Europe

and the Former Soviet Union can improve the regulatory quality (and quantity)

of their anti-corruption instructions, orders, and decrees.40 Regulations should

create social value instead of simply adding ―busy work.‖ Fewer police,

customs officers, doctors and other executive agency staff will take bribes if

they are exposed to greater legal liability and if they have other financial

deterrents. Executive agencies in countries like Russia, Bulgaria, and Armenia–

and the other countries in the region–should take greater jurisdiction over non-

criminal corruption cases. Keeping the processing ―in-house‖ can significantly

lower costs and raise financial collections (as well as mitigate social harms) of

corruption affecting these agencies. These agencies should be allocated budgets

to fight corruption–not based on the amount of money they received the

previous year–but based on their proven efficiency at ―catching bad guys‖ (or

preventing incentives from arising that creates these bad guys in the first place).

Qui tam rewards offer one example of a mechanism that can use economic

incentives to dissuade bribe-taking by customs officers, traffic police and other

members of executive agencies. While the 1990s and 2000s represented a

40. Such a view hopes to make a contribution to the literature by diverting attention away from strictly legislative changes. For an excellent analysis of such changes occurring in the Eastern European and Former Soviet region, see Beverley Earle, Bribery and Corruption in Eastern Europe, the Baltic States, and the Commonwealth of Independent States: What is to be Done?, 33 CORNELL INT’L L.J. 483, 492-509 (2000).


beginning for anti-corruption regulation, many countries in the region will

need to do much more work to adopt the instructions, decrees, orders and other

regulations needed to implement national anti-corruption laws and actually

fight corruption.

issues in anti-corruption law - university of notre …ndlaw/jleg/issue_articles/volume3… · ·...

Documents